Prompt Engineering Guide
Mastering Meeting notes extraction
on Grok-1
Stop guessing. See how professional prompt engineering transforms Grok-1's output for specific technical tasks.
The "Vibe" Prompt
"Extract the key decisions, action items, and follow-up tasks from the following meeting transcript:
[TRANSCRIPT]
"
Low specificity, inconsistent output
Optimized Version
You are a highly efficient meeting assistant with a focus on precision and conciseness. Your task is to extract critical information from meeting transcripts. Follow these steps:
1. **Identify Key Decisions:** Scan the transcript for explicit statements of decisions made, agreements reached, or conclusions drawn. List them clearly.
2. **Extract Action Items:** Look for specific tasks assigned to individuals or teams, including what needs to be done, by whom, and if a deadline is mentioned, include it. If no specific owner is mentioned but a task is clearly defined, list it as a general action item.
3. **Note Follow-up Tasks/Discussion Points:** Identify any items that require further discussion, research, or are slated for a future meeting. These are not necessarily immediate action items but topics that need to be revisited or explored.
4. **Synthesize and Format:** Present the extracted information clearly under the headings 'Key Decisions', 'Action Items', and 'Follow-up/Discussion Points'. Use bullet points for each entry.
***
Meeting Transcript:
[TRANSCRIPT]
Extracted Information:
Structured, task-focused, reduced hallucinations
Engineering Rationale
The optimized prompt provides a clear, step-by-step instruction set (chain-of-thought) for Grok-1. It defines specific categories to look for, guiding the model's extraction process. This structured approach reduces ambiguity, improves extraction accuracy, and directly prompts for a formatted output, minimizing the need for the model to 'guess' the desired structure or content. The persona ('highly efficient meeting assistant') also subtly reinforces the desired output quality.
0%
Token Efficiency Gain
The optimized prompt explicitly defines 'Key Decisions', 'Action Items', and 'Follow-up Tasks', leading to more accurate categorization.
The step-by-step guidance ('chain-of-thought') significantly reduces the model's hallucination rate for extracted information.
The output format specified in the optimized prompt ensures consistent and easy-to-read meeting notes.
Ready to stop burning tokens?
Join 5,000+ developers using Prompt Optimizer to slash costs and boost LLM reliability.
Optimize My Prompts